Evaluating sampling efficiency in depletion surveys using hierarchical Bayes

Abstract

Estimating animal abundance is essential to natural resource management and conservation. However, the cost associated with abundance estimation can be high for populations that are difficult to sample. Researchers, particularly in fisheries management, often sample such populations using depletion or removal surveys. Depletion surveys rely upon successive removals of animals, without replacement, to estimate abundance. These researchers also must decide on other sampling protocol, including the depletion technique, which may include depletion gear-type, vessel, or personnel. To inform this decision, we propose a supplement to the hierarchical Bayesian models recently introduced for the analysis of depletion data. Using Bayesian sample size methodology along with hierarchical modeling, we present a method for estimating the efficiency of previously employed depletion techniques. Using this method, the researcher can estimate the expected variability in abundance estimates for each depletion technique and apply this information to future decisions. Additionally, this method allows the estimation of expected variability for various numbers of depletion passes. We demonstrate the methodology using a data set of Chesapeake Bay blue crab ( Callinectes sapidus ) depletion surveys.